BACKGROUND Dobutamine stress echocardiography (DSE) is sensitive and specific in detecting myocardial ischemia of male patients. However, there have been few reports about the use of DSE for the detection of coronary artery disease (CAD) in women. METHODS DSE was evaluated in 51 consecutive women who underwent concomitant quantitative coronary angiography. Forty-four of the 51 patients received stress thallium-201 single-photon emission computed tomography (SPECT), and 30 of the 51 patients had interpretable results (exercise level > or = 85% of age-predicted maximal heart rate) of treadmill exercise. Twenty-nine patients had angiographically documented CAD defined as > or = 50% diameter stenosis. RESULTS The overall sensitivity of DSE and stress 201Tl SPECT in detecting CAD was 93% and 79% (p = nonsignificant), and the specificity was 82% and 75% (p = nonsignificant), respectively. A combination of both tests increased the sensitivity (96%) at the expense of some decrease in specificity (60%). The agreement of DSE and 201Tl SPECT was 68% (30 of 44; kappa statistic = 0.35; p < 0.0001). The overall sensitivity, specificity, and accuracy in detecting CAD by treadmill exercise test and DSE were 71% vs 93% (p = nonsignificant), 44% vs 82% (p = 0.036), and 57% vs 88% (p = 0.003). In patients with abnormal results of treadmill exercise testing, the false-positive rate in detecting CAD was 2 (18%) of 11 in patients with abnormal results of DSE and 7 (88%) of 8 in those with normal results of DSE (p = 0.005). In patients with normal results of treadmill exercise testing, the false-negative rate in detecting CAD was 4 (100%) of 4 in patients with abnormal results of DSE and 0 (0%) of 7 in those with normal results of DSE (p = 0.003). CONCLUSIONS The diagnostic accuracy of DSE was similar to that of stress 201Tl SPECT in women. DSE was able to stratify female patients with either abnormal or normal results of treadmill exercise testing and to avoid unnecessary cardiac catheterization.